Review: The move towards increasing specialisations comes with advantages as well as limitations. We might have increasingly detailed knowledge but we can't see how it connects. In the past few years there has been a move back towards integration and an understanding of the impact of connectedness. This is not to say that there have been no calls for a more holistic approach - one only has to think of Lovelock's Gaia to appreciate this. What is noticeable is the more focussed examination of interactions with a specific emphasis on social-environmental linkages. This is not the 1970s version with a desire for world peace but a far more measured approach, often model driven to consider the resilience of systems. This last idea ia crucial - we seem to have gone beyond the idea of simple interactions (which served well as a research focus for many years) to the application of interactions and the usefulness this might have for our futures. No doubt much of this is driven by climate modelling and the current emphasis on adaptation rather than mitigation but then it could be argued that its a win-win situation.

This text is the latest offering in this area and its origins are worth noting. It is the product of the Dahlem Workshops, named after the Berlin area they were held in. The aim is to get pre-organised focus on a specific range of topics so that the conference, when held, is a brainstorming session aimed at consolidating current expertise. Specific contributors are chosen with emphasis on a wide range of skills and disciplines since the key of these workshops is the promotion of inter-disciplinary research. It's also interesting to note that the lead editor is one whose work has been tied closely to the ecosystem service debate because many of these contributions are about the loss of functionality in ecosystems and the human impact this has.

The central argument of this text is that human-environment interactions occur at a range of timescales and that their overall impact is similarly diverse. The aim is to investigate those aspects which seem universal or dimensionless and so draw conclusions about how we affect the environment. To this end, the book is divided into five sections. The first, comprising three chapters, acts as an introduction providing the reader firstly with an overview of the book and its aims and then a critique of the quality of information in terms of models and data gathering. Section two tackles the millennial timescale. The focus is on the human-environment interactions that occurred up to 10,000 years ago. The classic case of the Maya is revisited along with other studies that together cover different societies, times and areas. The collapse of the Roman Empire due in part to environmental change is set alongside the development of Aboriginal Australia and the current controversy about the relative times and impacts of human and 'natural' change. In some ways, this chapter is a microcosm of the arguments presented here because each of the 'environment' and 'human' camps for change agents have their supporters but ultimately, data seem too poor to support either view alone (currently, there's a third suggesting both agents were to blame!). A chapter dealing with general human social decline theories brings us to a crucial and final chapter, the group report which argues, not surprisingly given the timescales and evidence, that there is no clear pathway through the debate but that it is almost certainly more complex than current ideas suggest. If millennia seem to long then section three looks at the century scale with a 1000 year span. Of course, it's the last 1000 years that have seen such a drastic change from an basically feudal peasant society to a globalised industrial one. The examples chosen highlight the impact of the environment on societies that could not always cope with the results. The first example is the less known El Niño of the late 18thC and its effect on colonial settlement. Given that the early Australian colony very nearly failed and that standard explanations fail to mention the weather there is still much that needs to be done, not least in publicising some events! Another example shows the impact of the opposite side of the coin - colder and wetter. The Little Ice Age was a complex event that impacted upon both social and environmental aspects. The lesson here, unlike the Australian case, is that mitigation and adaptation, even in rural societies, was both possible and likely. Dovetailed around these two chapters are others examining a very different set of cases. One, linking the modern nation-state, climate change, multinational corporations and civil protest, provides a highly original explanation of the current micropolitical scene although the conclusions are none too cheerful. The other reverses the past-to-present analytical viewpoint to examine the developments of the past from current ideas, notably information technology. In this section, the final report focusses on those factors which can be seen to be important at this scale. Whilst there is more that separates than combines the importance of history and the belief in the richness of resources from which we can draw did unite the contributors. Section four covers the shortest time span - the decade. Two chapters complement each other with their analyses of natural and human changes in the 20thC. A third chapter describes what can happen when modern social systems work in changing environments (in this case, desertification). Here, the group report acknowledges the great growth in all areas of human activity - population, resource consumption etc. since 1950. The final section looks towards the future. Since all predictions are no more than guesses the aim is to get the best quality data tools to make the guess the most accurate possible. Of the range that could be employed, three are chosen - scenario-making, global models and reflective analysis. This last one might surprise but it's based on a study of Limits to Growth and the stance is not so much to critique the book (that's probably happened enough) as to critique the responses. The aim here is clear - if the message is shot along with the messenger we need a better message.

This is a remarkable book on many levels. As an analysis it is original, both provoking and demanding of the reader with densely packed argument and detail that requires a god deal of prior knowledge and experience on a global scale. As an example of how we can learn from the past it highlights a range of case studies which complement work already done and extends our understanding. Finally, although this is not often said of an advanced science text ,it's a fascinating read pulling the reader through its arguments and data and demanding of some response. It's a pity that this book is too advanced for the senior student but it really should be a set text on undergraduate environmental and ecological courses as well as essential reading on teaching courses. Overall, one of the most remarkable texts published this year - make sure your library has a copy.